sábado, 13 de outubro de 2007

ROLAND BARTHES - Rhetoric of the Image

According to an ancient etymology, the word image should be linked to the root imitari. Thus we find ourselves im­mediately at the heart of the most important problem facing the semiology of images: can analogical representa­tion (the 'copy') produce true systems of signs and not merely simple agglutinations of symbols? Is it possible to conceive of an analogical 'code' (as opposed to a digital one)? We know that linguists refuse the status of language to all communication by analogy - from the 'language' of the 'language' of gesture - the moment such communications are not doubly articulated, a bees to re not founded on a combinatory system of digital units as phonemes are. Nor are linguists the only ones to be suspicious as to the linguis­tic nature of the image; general opinion too has a vague conception of the image as an are a of resistance to meaning ­this in the name of a certain mythical idea of Life: the image is re-presentation, which is to say ultimately resurrection, and, as we know, the intelligible is reputed antipathetic to lived experience. Thus from both sides the image is felt to be weak in respect of meaning: there are those who think that the image is an extremely rudimentary system in comparison with language and those who think that signi­fication cannot exhaust the image's ineffable richness. Now even - and above all if - the image is in a certain manner the limit of meaning, it permits the consideration of a veritable ontology of the process of signification. How does meaning get into the image? Where does it end? And if it ends, what is there beyond? Such are the questions that I wish to raise by submitting the image to a spectral analysis of the messages it may contain. We will start by making it considerably easier for ourselves: we will only study the advertising image. Why? Because in advertising the signification of the image is undoubtedly intentional; the signifieds of the advertising message are formed a priori by certain attributes of the product and these signifieds have to be transmitted as c1ear1y as possible. If the image contains signs, we can be sure that in advertising these signs are full, formed with a view to the optimum reading: the advertising image is frank, or at least emphatic.

The three messages

Here we have a Panzani advertisement: some packets of pasta, a tin, a sachet, some tomatoes, onions, peppers, a mushroom, all emerging from a half-open string bag, in yellows and greens on a red background. Let us try to 'skim off' the different messages it contains.

The image immediately yields a first message whose substance is linguistic; its supports are the caption, which is marginal, and the labels, these being inserted into the natural disposition of the scene, 'en abyme'. The code from which this message has been taken is none other than that of the French language; the only knowledge required to deci­pher it is a knowledge of writing and French. In fact, this message can itself be further broken down, for the sign Panzani gives not simply the name of the firm but also, by its assonance, an additional signified, that of 'Italianicity'. The linguistic message is thus twofold (at least in this particular image): denotational and connotational. Since, however, we have here only a single typical sign, namely that of articulated (written) language, it will be counted as one message.

Putting aside the linguistic message, we are left with the pure image (even if the labels are part of it, anecdotally). This image straightaway provides a series of discontinuous signs. First (the order is unimportant as these signs are not linear), the idea that what we have in the scene represented is a return from the market. A signified which itself implies two euphoric values: that of the freshness of the products and that of the essentially domestic preparation for which they are destined. Its signifier is the half-open bag which lets the provisions spill out over the table, 'unpacked'. To read this first sign requires only a knowledge which is in some sort implanted as part of the habits of a very widespread culture where 'shopping around for oneself' is opposed to the hasty stocking up (preserves, refrigerators) of a more 'mechanical' civilization. A second sign is more or less equally evident; its signifier is the bringing together of the tomato, the pepper and the tricoloured hues (yellow, green, red) of the poster; its signified is Italy or rather Italianicity. This sign stands in a relation of redundancy with the connoted sign of the linguistic message (the Italian assonance of the name Panzani) and the knowledge it draws upon is already more particular; it is a specifically 'French' knowledge (an Italian would barely perceive the connotation of the name, no more probably than he would the Italianicity of tomato and pepper), based on a familiarity with certain tourist stereotypes. Continuing to explore the image (which is not to say that it is not entirely clear at the first glance), there is no difficulty in discovering at least two other signs: in the first, the serried collection of different objects transmits the idea of a total culinary service, on the one hand as though Panzani furnished everything necessary for a carefully balanced dish and on the other as though the concentrate in the tin were equivalent to the natural produce surrounding it; in the other sign, the composition of the image, evoking the memory of innumerable alimentary paintings, sends us to an aesthetic signified: the 'nature morte' or, as it is better expressed in other languages, the 'still life'; the knowledge on which this sign depends is heavily cultural. It might be suggested that, in addition to these four signs, there is a further information pointer, that which tells us that this is an advertisement and which arises both from the place of the image in the magazine and from the emphasis of the labels (not to mention the caption). This last information, however, is co-extensive with the scene; it eludes signification insofar as the advertising nature of the image is essentially functional: to utter some­thing is not necessarily to declare I am speaking, except in a deliberately reflexive system such as literature.

Thus there are four signs for this image and we will assume that they form a coherent whole (for they are all discontinuous), require a generally cultural knowledge, and refer back to signifieds each of which is global (for example, Italianicity), imbued with euphoric values. After the linguistic message, then, we can see a second, iconic message. Is that the end? If all these signs are removed from the image, we are still left with a certain informational matter; deprived of all knowledge, I continue to 'read' the image, to 'understand' that it assembles in a common space a number of identifiable (nameable) objects, not merely shapes and colours. The signifieds of this third message are constituted by the real objects in the scene, the signifiers by these same objects photographed, for, given that the relation between thing signified and image signifying in analogical representation is not 'arbitrary' (as it is in language), it is no longer necessary to dose the relay with a third term in the guise of the psychic image of the object. What defines the third message is precisely that the relation between signified and signifier is quasi-tautological; no doubt the photograph involves a certain arrangement of the scene (framing, reduction, flattening) but this transition is not a transformation (in the way a coding can be); we have here a loss of the equivalence characteristic of true sign systems and a statement of quasi-identity. In other words, the sign of this message is not drawn from an institutional stock, is not coded, and we are brought up against the paradox (to which we will return) of a message without a code. This peculiarity can be seen again at the level of the knowledge invested in the reading of the message; in order to 'read' this last (or first) level of the image, all that is needed is the knowledge bound up with our perception. That knowledge is not nil, for we need to know what an image is (children only learn this at about the age of four) and what a tomato, a string-bag, a packet of pasta are, but it is a matter of an almost anthropological knowledge. This message corresponds, as it were, to the letter of the image and we can agree to call it the literal message, as opposed to the previous symbolic message.

If our reading is satisfactory, the photograph analysed offers us three messages: a linguistic message, a coded iconic message, and a non-coded iconic message. The linguistic message can be readily separated from the other two, but since the latter share the same (iconic) substance, to what extent have we the right to separate them? It is certain that the distinction between the two iconic messages is not made spontaneously in ordinary reading: the viewer of the image receives at one and the same time the perceptual message and the cultural message, and it will be seen later that this confusion in reading corresponds to the function of the mass image (our concern here). The distinction, however, has an operational validity, analogous to that which allows the distinction in the linguistic sign of a signifier and a signified (even though in reality no one is able to separate the 'word' from - its meaning except by recourse to the metalanguage of a definition). If the distinc­tion permits us to describe the structure of the image in a simple and coherent fashion and if this description paves the way for an explanation of the role of the image in society, we will take it to be justified. The task now is thus to recon­sider each type of message so as to explore it in its generality, without losing sight of our aim of understanding the overall structure of the image, the final inter-relationship of the three messages. Given that what is in question is not a 'naive' analysis but a structural description,1 the order of the messages will be modified a little by the inversion of the cultural message and the literal message; of the two iconic messages, the first is in some sort imprinted on the second: the literal message appears as the support of the 'symbolic' message. Hence, knowing that a system which takes over the signs of another system in order to make them its signifiers is a system of connotation, we may say immediately that the literal image is denoted and the symbolic image connoted. Successively, then, we shall look at the linguistic message, the denoted image, and the connoted image.

The linguistic message

Is the linguistic message constant? Is there always textual matter in, under, or around the image? In order to find images given without words, it is doubtless necessary to go back to partially illiterate societies, to a sort of pictographic state of the image. From the moment of the appearance of the book, the linking of text and image is frequent, though it seems to have been little studied from a structural point or view. What is the signifying structure of 'illustration'? Does the image duplicate certain of the informations given in the text by a phenomenon of redundancy or does the text add a fresh information to the image? The problem could be posed historically as regards the classical period with its passion for books with pictures (it was inconceiv­able in the eighteenth century that editions of La Fontaine's. Fables should not be illustrated) and its authors such as Menestrier who concerned themselves with the relations between figure and discourse. Today, at the level of mass communications, it appears that the linguistic message is indeed present in every image: as title, caption, accompany­ing press article, film dialogue, comic strip balloon. Which shows that it is not very accurate to talk of a civilization of the image - we are still, and more than ever, a civiliza­tion of writing, writing and speech continuing to be the full terms of the informational structure. In fact, it is simply the presence of the linguistic message that counts, for neither its position nor its length seem to be pertinent (a long text may only comprise a single global signified, thanks to connotation, and it is this signified which is put in relation with the image). What are the functions of the linguistic message with regard to the (twofold) iconic message? There appear to be two: anchorage and relay.

As will be seen more clearly in a moment, all images are polysemous; they imply, underlying their signifiers, a 'floating chain' of signifieds, the reader able to choose some and ignore others. Polysemy poses a question of meaning and this question always comes through as a dysfunction, even if this dysfunction is recuperated by society as a tragic (silent, God provides no possibility of choosing between signs) or a poetic (the panic 'shudder of meaning' of the Ancient Greeks) game; in the cinema itself, traumatic images are bound up with an uncertainty (an anxiety) concerning the meaning of objects or attitudes. Hence in every society various techniques are developed intended to fix the floating chain of signifieds in such a way as to counter the terror of uncertain signs; the linguistic message is one of these techniques. At the level of the literal message, the text replies - in a more or less direct, more or less partial manner - to the question: what is it? The text helps to identify purely and simply the elements of the scene and the scene itself; it is a matter of a denoted descrip­tion of the image (a description which is often incomplete) or, in Hjelmslev's terminology, of an operation (as opposed to connotation). The denominative function corresponds exactly to' an anchorage of all the possible (denoted) mean­ings of the object by recourse to a nomenclature. Shown a plateful of something (in an Amieux advertisement), may hesitate in identifying the forms and masses; the caption ('rice and tuna fish with mushrooms') helps me to choose the correct level of perception, permits me to focus not simply my gaze but also my understanding. When it comes to the 'symbolic message', the linguistic message no longer guides identification but interpretation, constituting a kind of vice which holds the connoted meanings from proliferating, whether towards excessively individual regions (it limits, that is to say, the projective power of the image) or towards dysphoric values. An advertisement (for d'Arcy preserves) shows a few fruits scattered around a ladder; the caption ('as if from your own garden') banishes one possible signified (parsimony, the paucity of the harvest) because of its un­pleasantness and orientates the reading towards a more flattering signified (the natural and personal character of fruit from a private garden); it acts here as a counter-taboo, combatting the disagreeable myth of the artificial usually associated with preserves. Of course, elsewhere than in ad­vertising, the anchorage may be ideological and indeed this is its principal function; the text directs the reader through the signifieds of the image, causing him to avoid some and receive others; by means of an often subtle dispatching, it remote-controls him towards a meaning chosen in advance. In all these cases of anchorage, language clearly has a function of elucidation, but this elucidation is selec­tive, a metalanguage applied not to the totality of the iconic message but only to certain of its signs. The text is indeed the creator's (and hence society's) right of inspection over the image; anchorage is a control, bearing a responsibility _ in the face of the projective power of pictures - for the use of the message. With respect to the liberty of the signifieds of the image, the text has thus a repressive value1 and we can see that it is at this level that the morality and ideology of a society are above all invested.

Anchorage is the most frequent function of the linguistic message and is commonly found in press photographs and advertisements. The function of relay is less common (at least as far as the fixed image is concerned); it can be seen particularly in cartoons and comic strips. Here text (most often a snatch of dialogue) and image stand in a complementary relationship; the words, in the same way as the images, are fragments of a more general syntagm and the unity of the message is realized at a higher level, that of the story, the anecdote, the diegesis (which is ample confirma­tion that the diegesis must be treated as an autonomous system). While rare in the fixed image, this relay-text becomes very important in film, where dialogue functions not simply as elucidation but really does advance the action by setting out, in the sequence of messages, meanings that are not to be found in the image itself. Obviously, the two functions of the linguistic message can co-exist in the one iconic whole, but the dominance of the one or the other is of consequence for the general economy of a work. When the text has the diegetic value of relay, the information is more costly, requiring as it does the learning of a digital code (the system of language); when it has a substitute value (anchorage, control), it is the image which detains the informational charge and, the image being analogical, the information is then 'lazier': in certain comic strips intended for 'quick' reading the diegesis is confided above all to the text, the image gathering the attributive informa­tions of a paradigmatic order (the stereotyped status of the characters); the costly message and the discursive message are made to coincide so that the hurried reader may be spared the boredom of verbal 'descriptions', which are entrusted to the image, that is to say to a less 'laborious' system.
We have seen that in the image properly speaking, the dis­tinction between the literal message and the symbolic message is operational; we never encounter (at least in advertising) a literal image in a pure state. Even if a totally 'naive' image were to be achieved, it would immediately join the sign of naivety and be completed by a third ­symbolic - message. Thus the characteristics of the literal message cannot be substantial but only relational. It is first of all, so to speak, a message by eviction, constituted by what is left in the image when the signs of connotation are mentally deleted (it would not be possible actually to remove them for they can impregnate the whole of the image, as in the case of the 'still life composition'). This evictive state naturally corresponds to a plenitude of virtualities: it is an absence of meaning full of all the meanings. Then again (and there is no contradiction with what has just been said), it is a sufficient message, since it has at least one meaning at the level of the identification of the scene represented; the letter of the image corresponds in short to the first degree of intelligibility (below which the reader would perceive only lines, forms, and colours), but this intelligi­bility remains virtual by reason of its very poverty, for everyone from a real society always disposes of a knowledge superior to the merely anthropological and perceives more than just the letter. Since it is both evictive and sufficient, it will be understood that from an aesthetic point of view the denoted image can appear as a kind of Edenic state of the image; cleared utopianically of its connotations, the image would become radically objective, or, in the last analysis, innocent.
This utopian character of denotation is considerably reinforced by the paradox already mentioned, that the photograph (in its literal state), by virtue of its absolutely analogical nature, seems to constitute a message without a code. Here, however, structural analysis must differentiate, for of all the kinds of image only the photograph is able to transmit the (literal) information without forming it by means of discontinuous signs and rules of transformation. The photograph, message without a code, must thus be opposed to the drawing which, even when denoted, is a coded message. The coded nature of the drawing can be seen at three levels. Firstly, to reproduce an object or a scene in a drawing requires a set of
rule-governed trans­positions; there is no essential nature of the pictorial copy and the codes of transposition are historical (notably those concerning perspective). Secondly" the operation of the drawing (the coding) immediately necessitates a certain divi­sion between the significant and the insignificant: the draw­ing does not reproduce everything (often it reproduces very little), without its ceasing, however, to be a strong message; whereas the photograph, although it can choose its subject, its point of view and its angle, cannot intervene within the object (except by trick effects). In other words, the denota­tion of the drawing is less pure than that of the photo­graph, for there is no drawing without style. Finally, like all codes, the drawing demands an apprenticeship (Saussure attributed a great importance to this semiological fact). Does the coding of the denoted message have consequences for the connoted message? It is certain that the coding of the literal prepares and facilitates connotation since it at once establishes a certain discontinuity in the image: the 'execution' of a drawing itself constitutes a connotation. But at the same time, insofar as the drawing displays its coding, the relationship between the two messages is profoundly modified: it is no longer the relationship between a nature and a culture (as with the photograph) but that between two cultures; the 'ethic' of the drawing is not the same as that of the photograph.

In the photograph - at least at the level of the literal message - the relationship of signifieds to signifiers is not one of 'transformation' but of 'recording', and the absence of a code clearly reinforces the myth of photographic 'naturalness': the scene is there, captured mechanically, not humanly (the mechanical is here a guarantee of objectivity). Man's interventions in the photograph (framing, distance, lighting, focus, speed) all effectively belong to the plane of connotation; it is as though in the beginning (even if utopian) there were a brute photograph (frontal and clear) on which man would then lay out, with the aid of various techniques, the signs drawn from a cultural code. Only the opposition of the cultural code and the natural non-code can, it seems, account for the specific character of the photograph and allow the assessment of the anthropological revolution it represents in man's history. The type of consciousness the photograph involves is indeed truly unprecedented, since it establishes not a consciousness of the being-there of the thing (which any copy could provoke) but an awareness of its having-been-there. What we have is a new space-time category: spatial immediacy and temporal anteriority, the photograph being an illogical conjunction between the here-now and the there-then. It is thus at the level of this denoted message or message without code that the real unreality of the photograph can be fully understood: its unreality is that of the here-now, for the photograph is never experienced as illusion, is in no way a presence (claims as to the magical character of the photographic image must be deflated); its reality that of the having-been-there, for in every photograph there is the always stupefying evidence of this is how it was, giving us, by a precious miracle, a reality from which we are sheltered. This kind of temporal equili­brium (having-been-there) probably diminishes the projec­tive power of the image (very few psychological tests resort to photographs while many use drawings): the this was so easily defeats the it's me. If these remarks are at all correct, the photograph must be related to apure spectatorial consciousness and not to the more projective, more 'magical' fictional consciousness on which film by and large depends. This would lend authority to the view that the distinction be­tween film and photograph is not a simple difference of degree but a radical opposition. Film can no longer be seen as animated photographs: the having-been-there gives way before a being-there of the thing; which omission would explain how there can be a history of the cinema, without any real break with the previous arts of fiction, whereas the photograph can in some sense elude history (despite the evolution of the techniques and ambitions of the photographic art) and represent a 'flat' anthropological fact, at once absolutely new and definitively unsurpassable, humanity encountering for the first time in its history messages without a code. Hence the photograph is not the last (improved) term of the great family of images; it corresponds to a decisive mutation of informational economies.

At all events, the denoted image, to the extent to which it does not: imply any code (the case with the advertising photograph), plays a special role in the general structure of the iconic message which we can begin to define (returning to this question after discussion of the third message): the denoted image naturalizes the symbolic message, it innocents the semantic artifice of connotation, which is extremely dense, especially in advertising. Although the Panzani poster is full of 'symbols', there nonetheless remains in the photograph, insofar as the literal message is suffi­cient, a kind of natural being-there of objects: nature seems spontaneously to produce the scene represented. A pseudo-truth is surreptitiously substituted for the simple validity of openly semantic systems; the absence of code disintelIec­tualizes the message because it seems to found in nature the signs of culture. This is without doubt an important historical paradox: the more technology develops the diffusion of information (and notably of images), the more it provides the means of masking the constructed meaning under the appearance of the given meaning.

Rhetoric of the image

It was seen that the signs of the third message (the 'symbolic' message, cultural or connoted) were discontinuous. Even when the signifier seems to extend over the whole image, it is nonetheless a sign separated from the others: the 'composition' carries an aesthetic signified, in much the same way as intonation although suprasegmental is a separate signifier in language. Thus we are here dealing with a normal system whose signs are drawn from a cultural code (even if the linking together of the elements of the sign appears more or less analogical). What gives this system its originality is that the number of readings of the same lexical unit or lexia (of the same image) varies according to individuals. In the Panzani advertisement analysed, four connotative signs have been identified; probably there are others (the net bag, for example, can signify the miraculous draught of fishes, plenty, etc.). The variation in readings is not, however, anarchic; it depends on the different kinds of knowledge - practical, national, cultural, aesthetic - invested in the image and these can be classified, brought into a typology. It is as though the image presented itself to the reading of several different people who can perfectly well co-exist in a single individual: the one lexia mobilizes different lexicons. What is a lexicon? A portion of the symbolic plane (of language) which corresponds to a body of practices and techniques. This is the case for the different readings of the image: each sign corresponds to a body of 'attitudes' - tourism, housekeeping, knowledge of art ­certain of which may obviously be lacking in this or that individual. There is a plurality and a. co-existence of lexicons in one and the same person, the number and identity of these lexicons forming in some sort a person's idiolect. The image, in its connotation, is thus constituted by an architecture of signs drawn from a variable depth of lexicons (of idiolects); each lexicon, no matter how 'deep', still being coded, if, as is thought today, the psyche itself is articulated like a language; indeed, the further one 'descends' into the psychic depths of an individual, the more rarified and the more classifiable the signs become - what could be more systematic than the readings of Rorschach tests? The variability of readings, therefore, is no threat to the 'language' of the image if it be admitted that that language is composed of idiolects, lexicons and sub-codes. The image is penetrated through and through by the system of meaning, in exactly the same way as man is articulated to the very depths of his being in distinct languages. The language of the image is not merely the totality of utterances emitted (for example at the level of the combiner of the signs or creator of the message), it is also the totality of utterances received: the language must include the 'surprises' of meaning.

Another difficulty in analysing connotation is that there is no particular analytical language corresponding to the particularity of its signifieds - how are the signifieds of connotation to be named? For one of them we ventured the term Italianicity, but the others can only be designated by words from ordinary language (culinary preparation, still life, plenty); the metalanguage which has to take charge of them at the moment of the analysis is not specialized. This is a difficulty, for these signifieds have a particular semantic nature; as a seme of connotation, 'plenty' does not exactly cover 'plenty' in the denoted sense; the signifier of connotation (here the profusion and the condensation of the produce) is like the essential cipher of all possible plenties, of the purest idea of plenty. The denoted word never refers to an essence for it is always caught up in a contingent utterance, a continuous syntagm (that of verbal discourse), oriented towards a certain practical transitivity of language; the seme 'plenty', on the contrary, is a concept in a pure state, cut off from any syntagm, deprived of any context and corresponding to a sort of theatrical state of meaning, or, better (since it is a question of a sign without a syntagm), to an exposed meaning. To express these semes of connota­tion would therefore require a special metalanguage and we are left with barbarisms of the Italianicity kind as best being able to account for the signifieds of connotation, the suffix -icity deriving an abstract noun from the adjective: Italianicity is not Italy, it is the condensed essence of every­thing that could be Italian, from spaghetti to painting. By accepting to regulate artificially - and if needs be barbarously - the naming of the semes of connotation, the analysis of their form will be rendered easier. These semes are organized in associative fields, in paradigmatic articulations, even perhaps in oppositions, according to certain defined paths or, as A. J. Greimas puts it, according to certain semic axes: Italianicit y belongs to a certain axis of nationalities, alongside Frenchicity, Germanicity or Spanishicity. The reconstitution of such axes - which may eventually be in opposition to one another - will clearly only be possible once a massive inventory of the systems of connotation has been carried out, an inventory not merely of the connotative system of the image but also of those of other substances, for if connotation has typical signifiers dependent on the different substances utilized (image, language, objects, modes of behaviour) it holds all its signifieds in common: the same signifieds are to be found in the written press, the image or the actor's gestures (which is why semiology can only be conceived in a so to speak total framework). This common domain of the signifieds of connotation 'is that of ideology, which cannot but be single for a given society and history, no matter what signifiers of connotation may use.

To the general ideology, that is, correspond signifiers of connotation which are specified according to the chosen substance. These signifiers will be called connotators and the set of connotators a rhetoric, rhetoric thus appearing as the signifying aspect of ideology. Rhetorics inevitably vary by their substance (here articulated sound, there image, gesture 'or wI1atever) but not necessarily by their form; it is even probable that there exists a single rhetorical form, common for instance to dream, literature and image. Thus the rhetoric of the image (that is to say, the classifica­tion of its connotators) is specific to the extent that it is subject to the physical constraints of vision (different, for example, from phonatory constraints) but general to the extent that the 'figures' are never more than formal rela­tions of elements. This rhetoric could only be established on the basis of a quite considerable inventory, but it is possible now to foresee that one will find in it some of the figures formerly identified by the Ancients and the Classics; the tomato, for example, signifies Italianicity by metonymy and in another advertisement the sequence of three scenes (coffee in beans, coffee in powder, coffee sipped in the cup) releases a certain logical relationship in the same way as an asyndeton. It is probable indeed that among the meta-bolas (or figures of the substitution of one signifier for another), it is metonymy which furnishes the image with the greatest number of its connotators, and that among the parataxes (or syntagmatic figures), it is asyndeton which predominates.

The most important thing, however, at least for the moment, is not to inventorize the connotators but to understand that in the total image they constitute dis­continuous or better still scattered traits. The connotators do not fill the whole of the lexia, reading them does not exhaust it. In other words (and this would be a valid proposition for semiology in general), not all the elements of the lexia can be transformed into connotators; there always remaining in the discourse a certain denotation without which, precisely, the discourse would not be possible. Which brings us back to the second message or denoted image. In the Panzani advertisement, the Mediterranean vegetables, the colour, the composition, the very profusion rise up as so many scattered blocks, at once isolated and mounted in a general scene which has its own space and, as was seen, its 'meaning': they are 'set' in a syntagm which is not their s and which is that of the denotation. This last proposition is important for it permits us to found (retro­actively) the structural distinction between the second or literal message and the third or symbolic message and to give a more exact description of the naturalizing function of the denotation with respect to the connotation. We can now understand that it is precisely the syntagm of the denoted message which 'naturalizes' the system of the connoted message. Or again: connotation is only system, can only be defined in paradigmatic terms; iconic denota­tion is only syntagm, associates elements without any system: the discontinuous connotators are connected, actualized, 'spoken' through the syntagm of the denotation, the discontinuous world of symbols plunges into the story of the denoted scene as though into a lustral bath of innocence.

It can thus be seen that in the total system of the image the structural functions are polarized: on the one hand there is a sort of paradigmatic condensation at the level of the connotators (that is, broadly speaking, of the symbols), which are strong signs, scattered, 'reified'; on the other a syntagmatic 'flow' at the level of the denotation - it will not be forgotten that the syntagm is always very close to speech, and it is indeed the iconic 'discourse' which natural­izes its symbols. Without wishing to infer too quickly from the image to semiology in general, one can nevertheless venture that the world of total meaning is torn internally (structurally) between the system as culture and the syn­tagm as nature: the works of mass communications all combine, through diverse and diversely successful dialectics, the fascination of a nature, that of story, diegesis, syntagm, and the intelligibility of a culture, withdrawn into a few discontinuous symbols which men 'decline' in the shelter of their living speech.